PurposeThe purpose of our study is to investigate the effect of the dividend tax reform established by the Job Creation Act on corporate dividend behavior in a manner that increases firms' dividends. We intend to test whether the rise in dividends following the new tax law is motivated by tax considerations.Design/methodology/approachUsing the data from the Indonesian Stock Exchange (IDX) database and Bloomberg, we employ the difference-in-difference approach as our main strategy.FindingsOur results are twofold. First, we find that firms pay higher dividends after the tax reform, despite facing sharp declines in financial performance caused by the pandemic crisis. Therefore, the dividend rise among Indonesian firms after the tax reform cannot be attributed to corporate profitability as suggested by several scholars who studied the effect of the dividend tax reform on corporate dividend behavior in the USA. Second, we find that firms pay dividends more frequently following the dividend tax reform. As a result, we confirm that our results are not driven by a few large payers, and there is a larger number of firms that are initiating or increasing dividends compared to those that are omitting or decreasing dividends.Research limitations/implicationsWe face two limitations in our paper. First, we do not separate the dividend payments into regular and special dividends as done by Chetty and Saez (2005) and Julio and Ikenberry (2004). Examining the response of the dividend tax reform separately between regular and special dividends allows for more valid inferences about whether the increase in dividend payments following the tax reform should be attributed to the change in firms' dividend behavior in general or the sole effect of a few large firms that paid one-time and non-recurring dividends that drove the results. Second, we provide evidence that may be particularly unique to developing countries like Indonesia, where controlling shareholders predominantly stir corporate dividend decisions. Consequently, it is challenging to generalize our findings to other contexts where the legal and regulatory environments are distinct.Practical implicationsFirst, our findings provide evidence that the dividend rise after the tax reform is not attributable to corporate profitability, which has long been claimed by several scholars as a factor that influences the dividend rise post-reform rather than the dividend policy per se. Second, our results demonstrate that, despite the fact that controlling shareholders lack marginal incentives to increase dividend payments, the dividend tax reform can serve as a governance mechanism to safeguard minority shareholders in the absence of strong investor protection. Lastly, our findings provide evidence that the dividend tax policy can become the first-order determinant that shapes firms' dividend decisions rather than firm-specific factors such as the pattern of past dividends and stability of earnings.Originality/valueThis is the first study that documents the effects of the reduction in tax penalty on corporate dividend behavior in Indonesia, where the dividend tax exemption is set to be permanently implemented.Table B1Univariate analysis between firms in the control and treatment groups following the dividend tax reformBeforeAfterDifferenceln(DivPayment)Control6.0463.5942.451** (1.132)Treatment5.1715.691-0.520* (0.343)Difference0.874 (0.845)-2.096*** (0.866)2.971*** (0.811)DiD Estimate with Covariates2.938*** (0.816)DivFrequencyControl0.2310.1370.093** (0.043)Treatment0.2070.229-0.021* (0.013)Difference0. 023 (0.033)-0.091*** (0.035)0.115*** (0.032)DiD Estimate with Covariates0.114*** (0.032)Note(s): Table 2 presents the univariate analysis of the effect of the dividend tax reform on corporate dividend behavior. The dependent variables are the logarithmic value of dividend payments and the binary outcome which takes a value of 1 if firms pay dividends in a given quarter and 0 otherwise. *, ** and *** indicate significance at the 10%, 5% and 1% levels, respectivelySource(s): Authors' own workOriginality/valueThis is the first study that documents the effects of the reduction in tax penalty on corporate dividend behavior in Indonesia, where the dividend tax exemption is set to be permanently implemented.Table B1Univariate analysis between firms in the control and treatment groups following the dividend tax reformBeforeAfterDifferenceln(DivPayment)Control6.0463.5942.451** (1.132)Treatment5.1715.691-0.520* (0.343)Difference0.874 (0.845)-2.096*** (0.866)2.971*** (0.811)DiD Estimate with Covariates2.938*** (0.816)DivFrequencyControl0.2310.1370.093** (0.043)Treatment0.2070.229-0.021* (0.013)Difference0.023 (0.033)-0.091*** (0.035)0.115*** (0.032)DiD Estimate with Covariates0.114*** (0.032)Note(s): Table 2 presents the univariate analysis of the effect of the dividend tax reform on corporate dividend behavior. The dependent variables are the logarithmic value of dividend payments and the binary outcome which takes a value of 1 if firms pay dividends in a given quarter and 0 otherwise. *, ** and *** indicate significance at the 10%, 5% and 1% levels, respectivelySource(s): Authors' own workOriginality/valueThis is the first study that documents the effects of the reduction in tax penalty on corporate dividend behavior in Indonesia, where the dividend tax exemption is set to be permanently implemented.Table B1Univariate analysis between firms in the control and treatment groups following the dividend tax reformBeforeAfterDifferenceln(DivPayment)Control6.0463.5942.451** (1.132)Treatment5.1715.691-0.520* (0.343)Difference0.874 (0.845)-2.096*** (0.866)2.971*** (0.811)DiD Estimate with Covariates2.938*** (0.816)DivFrequencyControl0.2310.1370.093** (0.043)Treatment0.2070.229-0.021* (0.013)Difference0.023 (0.033)-0.091*** (0.035)0.115*** (0.032)DiD Estimate with Covariates0.114*** (0.032)Note(s): Table 2 presents the univariate analysis of the effect of the dividend tax reform on corporate dividend behavior. The dependent variables are the logarithmic value of dividend payments and the binary outcome which takes a value of 1 if firms pay dividends in a given quarter and 0 otherwise. *, ** and *** indicate significance at the 10%, 5% and 1% levels, respectivelySource(s): Authors' own work
